3XS Cyclone RAM Fault

[Thomas_Asus]

Thanks for getting back to me with that information Thorbog, its very much appreciated. I have given all provided information back to our HQ.

May I ask, when you said "cold boot" are you referring to removing the AC power cable before booting up the machine? I say this because on X58 chipset motherboards the removal of the AC cable can in many cases affect RAM compatibility.

[Thorbog]

No the cable never gets removed.

COLD BOOT

By "Cold Boot" I mean turning the machine on when it is physically cold - i.e. it has been shut down for a few hours.

When cold booting it will fail to detect all the RAM and with the latest BIOS it will go into a BSOD/reboot loop.

I haven’t tested to find the minimum time the machine must be off to create “cold boot” conditions but I have seen it fail to detect all RAM after being shut down for as little as 20 mins.


WARM BOOT

By "Warm Boot" I mean booting up from having been recently shut down – i.e. just a few seconds/minutes after the machine was shut down.

When "warm booting" the machine works perfectly, detects all the RAM and the new BIOS works fine.

Note that “restarting” does not trigger a “warm boot”. It only works following a “shut down” where you will hear the fans stop. On a “restart” the fans and all the lights stay on and the system will continue to BSOD or fail to detect all the RAM. I guess the machine has to physically shut down to re-detect the RAM.

For the machine to be warm enough to trigger a “warm boot” it only has to have been on for maybe 30 seconds. So for example usually when I get home from work the machine will always boot with only 2GB or RAM. What I do though is turn the PC on and as soon as Windows loads immediately select shut down. After shutting down I immediately turn it on again and it will boot up nicely with all the RAM.

Then if I wanted I could shut down or restart a million times that evening and it will keep working perfectly……..so long as I don’t leave it shut down long enough for it to cool down…



I’ve been on a number of forums where there are epic threads on this issue and “cold boot” now seems to be an official recognised term.

[Thorbog]

One interesting thing about this latest BIOS...

I thought I would go into the BIOS on cold boot this evening before it started loading Windows and entering the BSOD cycle.

The BIOS was actually showing 6GB....

So it looks like that BIOS was detecting (or at least displaying that it had detected) all the RAM but sadly this is no good if it won't boot up fully!

[B.E.A.N]

Thorbog

Any results on your system yet?. I have had one half day when it is reporting ony 4GB since your last post (not counting this morning) so I am still loathe to try the beta fix and want to wait till Asus come up with a permanent solution to the problem. I just wonder if opting for the faster CMT6GX3M3A2000C8 Corsair GT changes the problem and thats why it shows the 4GB and only happens once in a while.
Hope you get your system working 100% soon.
john

[Vedeneyev]

OK just to report I'm still also getting this problem exactly as described. My PC shipped at the end of Oct with BIOS v0402, and it would more often than not report 10GB RAM instead of the installed 12GB.

After finding out it was a known problem and not something dumb I'd done, I flashed v0503, and during POST the board now always reported the full complement of 12GB, but from the get-go I got many BSOD cycles on the first boot of the day (with or without OC), with errors such as:
MEMORY_MANAGEMENT
PAGE_FAULT_IN_NONPAGED_AREA
PFN_LIST_CORRUPT

After tolerating several BSOD's, Windows startup repairs (with BSOD's) and system restores (with BSOD's), it would usually get to the point where Startup Repair would report it could not repair the system, but then on the next boot everything would be peachy.

That situation would last for the rest of the day until I did a full shutdown and left the PC for several hours (usually the next day-and without removing any power cables etc). Then we'd be back to square one on the first boot, spending about 20mins working through the BSOD's and repairs until it booted, after which everything would be fine no matter how many times I immediately shutdown (restart, or shutdown&start). Until of course the next 'cold' boot. Reinstalled Windows several times, ran MemTest over 7hrs with no faults, bought a new SSD so that I wasn't booting from RAID, all to no avail - the BSOD's were still the same.

So now I've tried v0506, zero-wiped my drives and done a fresh install of Windows, which showed some promise. Again at every POST it reports the full 12GB of RAM. Now I didn't get the BSOD's before booting into Windows, and most times I have a stable system. However occasionally, a few minutes after having booted into Windows doing nothing but email or web, the system crashes and either saying it's rebooting in 1min (and goes to BSOD) or just straight to BSOD with MEMORY_MANAGEMENT error.

SCAN/3XS and Asus have been brilliant at keeping me up to date and working toward a fix for this very peculiar problem, and I'm sure they'll get there in the end, but I'm sad to say I can't wait any longer and I think it's going to have to go back for a Gigabyte board...

[Thorbog]

There may be a quicker way rather than going through all those start up repairs and BSOD cycles.

When you first boot up each day, try going directly into the BIOS the first time you turn it on. Then when you are in the BIOS don't do anything, just turn the PC off by pressing the power button so that the fans stop etc.

Wait a second...then turn it on again.

That always worked for me and was the quickest way to boot up with all the RAM and no Blue Screens. I have a 100% success rate with that. Granted that isn't a permanent solution but it means you can boot up quickly with no messing around with errors.

Of course when I say that worked 100% for me, I only have 6GB. Could be different on a machine with 12GB.

Try that and see if you get any of those issues in Windows.

[Thomas_Asus]

Hi Vedeneyev,

May I ask, after you updated the BIOS to 0506 did you remember to clear the CMOS? With the alterations we are making this is an important step to make.

If/after you have cleared the CMOS and the instability issue is still present could you please try using the non-overclocked O.C. profile to see whether that yields different results.

Thank you.

[Vedeneyev]

Hi Thomas

After flashing to 0506, I did clear CMOS, loaded up with default settings, then reinput my overclock, then HDDErased all my drives and did a clean install of Win7.

I've just cleared CMOS again, loaded default BIOS settings, and only changed the HDD Setup so I can boot into Windows (2xRaptors on the Intel controller to RAID0 and 1xSSD on the Marvell to AHCI). First boot of the day today I get the following:

'Starting Windows' splash to blank black screen and locked up cursor/unresponsive keyboard
Hard Reboot
Startup Repair to BSOD with PFN_LIST_CORRUPT
Hard Reboot
Startup Repair to System Restore completes OK
Normal Reboot
System starts up normally, and I know it will now be OK for the rest of the day... until I come to boot tomorrow... even with default BIOS settings.

This is typical with what's been happening even with my unaggressive overclock (it's a 980X to only 3.6GHz with 200BCLK and 18x multi, 1.35V CPU and 1.65V for the 12GBx1600DDR3). I say it's typical but often the startup sequence has many more BSODs and Startup Repair attempts.

The problem's transient but repeatable, so I can't get my head round it? The only solution that works is Thorbog's to go straight into BIOS on the first boot of the day, then turn the sytem off and on without changing anything, then everything seems fine. I love the board and the BIOS when it's booted up, just wish it wasn't so painful starting each day. So any ideas please let me know!

[Vedeneyev]

Thomas,

Just to report I also get first-bootup-of-the-day problems booting into Linux (although again, not if I enter BIOS, shutdown and restart). When unsuccessful it simply hangs during the startup splash but will still allow CTRL-ALT-DEL to reboot.

I've had a look at the syslog's and compared between unsuccessful and successful boots, and although I'm no expert it seems to be pointing to something to do with the PCI slots and video cards. If you PM me your email address I can send you the logs if you think it might help?

[Thorbog]

Just to update you on my system.

We changed the mobo from Asus to Gigabyte and it made no difference. I had exactly the same blue screening on the first cold boot. So if you are getting ideas about changing your mobo I wouldn't bother.

So we replaced the CPU, RAM and PSU and that has fixed it! The machine now boots up first time from cold with no blue screen and a full compliment of memory!

Pity we don't know which of those 3 items was the cause, I dunno what Scan will do with them and if they can find which item was responsible.

Have to say though my problem was 100% consistent. Every time I cold booted it would blue screen unless I did the BIOS trick. It never successfully cold booted even once.

Are you saying yours sometimes works?

[Thomas_Asus]

Hi Thorbog,

I am happy to hear your system is now stable. May I please ask you what RAM you are now using in your system? Also what revision is it?

Many thanks.

[Vedeneyev]

[5th post so I can add URLs]

[Vedeneyev]

I have the exactly the same problem as you Thorbog, in that every time I boot I get Bluescreens going into Windows and lockups going into Linux (unless of course I do your BIOS trick).

Thomas, here are my syslogs from 3 failed boots followed by 2 successful boots after having done Thorbog's BIOS trick (It's booting from sdd5 and I have Corsair CMP6GX3M3A1600C7 ver7.1A)
Boot1fail
Boot2fail
Boot3fail
Boot4success
Boot5success

Let me know if it helps at all - replacing the MoBo, RAM, CPU & PSU sounds like a nuclear option!

[Winterlove]

I've had my fixed machine for a few weeks now. To recap, mine was fixed by having the motherboard replaced, and the memory and RAM was replaced too, so the original failing CPU-RAM-motherboard combination could go back to Asus for analysis.

I too initially experienced the cold boot problems. What would happen is that it would get someway into the POST, then fail, shut down and then restart with extremely conservative clock settings. If I then went into BIOS I could simply resave the OC setup and my normal overclock would boot up as normal.

Very rarely the failure mode was a freeze during the Windows boot process (if the Windows 7 swirly lights don't appear, I know there is a problem), but usually it is a failure to POST.

So, I reasoned that the machine *really* didn't like being started cold. Since my first boot of the day was usually early in the morning, when the house was cool, I tried just putting the radiator in the study on (I normally kept it off because it seems silly to heat a room with a several computers in it ...). Since doing that I think I've had one POST failure in two or three weeks, rather than every day.

So, I'm wondering if there are some thermal tolerance/flexing issues lurking here. However, I could be talking rubbish - it happens sometimes.

[PeteG@Scan]

Winterlove: Yes, I was going to contact you after we gave a solution a try on Thorbogs system. The voltages not holding at startup now was I'm guessing a failure of the psu to give a stable voltage right at start up on a cold start. I'm starting to see a pattern here with a few and I wanted to narrow it down a bit before we started to offer it as a solution. I think we may have it down but it'll be a few more days until were certain of it. Thorbogs still got a few issues now but I've got a feeling it's down to the overclock and a few settings needing tweaking now rather than a unsolvable problem.

I'll let you know as soon as we have an answer.

[RobCSQ]

Hi all

I purchased 2 identical systems using this Asus Mobo and this RAM issue is prevalent on one of them but the other has had no such issues which I find strange if this is regarded as a BIOS problem. I don't personally use the faulty system, one of our junior designers does, he has logged the fault through the online support page but we're still looking for a fix.

So far I have flashed the BIOS to 0405 (as this was the newest one I could find) on the advice of 3XS Support, but this has brought its own issues. Has anybody discovered a solution to this yet? From what I have been reading this is starting to seem more like a hardware fault than a BIOS one!?